Re: [PATCH v5 2/2] arm64: dts: qcom: sc7180-trogdor: Add lpass dai link for HDMI

[Doug Anderson]

Hi,


On Thu, Jul 22, 2021 at 7:49 AM Srinivasa Rao Mandadapu
<srivasam@xxxxxxxxxxxxxx> wrote:
>
> Thanks for your time Doug!!
>
> On 7/21/2021 9:50 PM, Doug Anderson wrote:
> > Hi,
> >
> > On Wed, Jul 21, 2021 at 1:42 AM Srinivasa Rao Mandadapu
> > <srivasam@xxxxxxxxxxxxxx> wrote:
> >> From: V Sujith Kumar Reddy <vsujithk@xxxxxxxxxxxxxx>
> >>
> >> Add dai link in sc7180-trogdor.dtsi for supporting audio over DP
> >>
> >> Signed-off-by: V Sujith Kumar Reddy <vsujithk@xxxxxxxxxxxxxx>
> >> Signed-off-by: Srinivasa Rao Mandadapu <srivasam@xxxxxxxxxxxxxx>
> >> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@xxxxxxxxxx>
> >> ---
> >>   arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi | 17 +++++++++++++++++
> >>   1 file changed, 17 insertions(+)
> >>
> >> diff --git a/arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi b/arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi
> >> index 31bf7c698b8f..a4cb9ee567ff 100644
> >> --- a/arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi
> >> +++ b/arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi
> >> @@ -288,6 +288,7 @@ sound: sound {
> >>                          "Headphone Jack", "HPOL",
> >>                          "Headphone Jack", "HPOR";
> >>
> >> +               #sound-dai-cells = <0>;
> > The `#sound-dai-cells` is not in the bindings. The bindings also say
> > `additionalProperties: false`. So either your patch is wrong for
> > including this or the bindings are wrong for omitting it. Which is it?
> > I notice that downstream we don't have this.
> >
> Yes it's wrongly placed Here. Will remove it.
> >> @@ -314,6 +315,18 @@ sound_multimedia1_codec: codec {
> >>                                  sound-dai = <&max98357a>;
> > Ideally you'll want to rebase to make context clean since the above
> > got changed from `max98357a` to `max98360a`.
> >
> >
> >>                          };
> >>                  };
> >> +
> >> +               dai-link@2 {
> >> +                       link-name = "MultiMedia2";
> >> +                       reg = <2>;
> > I am certainly not an expert, but I notice that downstream we have reg
> > as <LPASS_DP_RX>, which makes the node name dai-link@5 instead of @2.
> > Does that matter?
> >
> Yes,Here reg as <LPASS_DP_RX> is correct option. Will change
> accordingly.  But It seems dai-link@2 is appropriate as in this platform

I think you misunderstand. The "@num" is _required_ to match the value
for "reg". So if reg is 5 then you need "@5", not "@2"


> we are using only 3 nodes. i.e. I2S primary, I2S secondary and DP node.
>
> >> @@ -768,6 +781,10 @@ secondary_mi2s: mi2s@1 {
> >>                  reg = <MI2S_SECONDARY>;
> >>                  qcom,playback-sd-lines = <0>;
> >>          };
> >> +
> >> +       hdmi-primary@0 {
> >> +               reg = <LPASS_DP_RX>;
> > Your node name and `reg` don't match. `LPASS_DP_RX` is 5 so the node
> > name should include @5, not @0. I also notice that downstream the node
> > name is called `hdmi`, not `hdmi-primary`. The downstream `hdmi` seems
> > more parallel to the sibling nodes, like `mi2s@0` and `mi2s@1`.
> Yes hdmi@2 is appropriate. will change accordingly.

Same here. Since LPASS_DP_RX is actually the number 5 then you need
@5. I think there was discussion about allowing it to be @LPASS_DP_RX
but last I remember the decision was to keep it as a number.

-Doug